- Get link
- Other Apps
- Get link
- Other Apps
In a technology-intensive, greener and potentially deglobalizing world, local skills development will matter more than ever
The rise of new generative Artificial Intelligence (AI) systems has captured global attention with their immense potential. While AI's capacity to analyze vast amounts of data and make autonomous decisions is both impressive and unsettling, concerns have been raised about bias, privacy invasion, job displacement, and the fear of machines running amok. So, how do we ensure that AI benefits society?
In response, the National Telecommunications and Information Administration (NTIA) has solicited input on establishing "accountability" for AI companies. They are exploring ways to develop a productive AI accountability ecosystem. NTIA's definition of "AI" is broad enough to encompass most significant software systems, even pre-dating recent AI advancements.
However, as highlighted in our recently submitted comments at the Center for Growth and Opportunity, NTIA's broad approach overlooks the primary mechanism through which companies and technology are held accountable. Just like businesses in any industry, AI companies are primarily held accountable by their customers within a competitive market system. Business and consumer markets, reputational and financial markets, applicable laws, and societal norms collectively create feedback loops that align the interests of producers with stakeholders. Competition for profit drives this alignment, fostering innovation, the development of valuable products and services, and ultimately benefiting society.
Admittedly, this system is not flawless. Occasionally, market feedback mechanisms may fail, particularly when a producer's actions impact third parties or when there is asymmetrical information distribution. In such cases, alternative mechanisms may be necessary, but they should be exceptions rather than the norm, complementing market accountability mechanisms rather than replacing them.
NTIA's request fails to acknowledge market-based accountability or identify gaps in this context. Yet, many of the proposed AI accountability mechanisms, such as transparency, certifications, and third-party audits, can and already do function within the market, both in AI and other domains.
Consider, for instance, products that certify the integrity of their supply chains with seals, or the "UL" certification marks on various electrical home devices, or the Yelp ratings of local services and restaurants. In the realm of AI, a recent example is the Center for Industry Self-Regulation's release of the Principles for Trustworthy AI in Recruiting and Hiring, along with Independent Certification Protocols for AI-Enabled Hiring and Recruiting Technologies. These initiatives aim to establish global standards and provide a pathway for independent certification in the use of AI applications for recruitment and hiring. Market-based approaches like these are crucial for fostering a thriving AI accountability ecosystem.
It's important to note that while the NTIA plays a significant role in facilitating discussions, it lacks regulatory authority and cannot impose binding rules. Therefore, the agency cannot afford to overlook market-based accountability, given its limited mandate and the inherent complexity of AI. AI accountability goes beyond regulation; it involves engaging with our existing market system, identifying gaps, and implementing exceptions cautiously when necessary. It entails preserving the market's capacity to self-correct and innovate. Indeed, markets can help society navigate the inevitable trade-offs between different accountability goals, such as privacy and transparency, or accuracy and access.
NTIA's exploration of AI accountability should be more than a mere exercise in considering how governments or companies might regulate themselves. It should transcend hypothetical concerns about AI that may require regulation in the future. Instead, NTIA should adopt a holistic perspective, recognizing that accountability primarily stems from markets. The focus should be on enhancing these market-based accountability ecosystems in the era of AI, rather than attempting to replace or undermine them.
By prioritizing the strengthening of our oldest and most robust accountability mechanisms provided by markets, NTIA will be better equipped to address any potential harms while preserving the societal benefits of AI.
- Get link
- Other Apps
Comments
Post a Comment